The cold-climate hypothesis for evolution of viviparity in squamates predicts a correlation between reproductive mode, altitude and latitude. I tested this prediction in horned lizards within a phylogenetic context. I first determined whether all viviparous species were monophyletic using Monte Carlo simulations. Secondly, I tested for presence of phylogenetic signal using randomization tests. Thirdly, I analysed relationships between reproductive mode and minimum, midpoint, and maximum altitudes and latitudes by computing conventional correlations and phylogenetically independent contrasts. Viviparous species do not form a monophyletic group suggesting viviparity evolved twice in the genus. Viviparity and altitude showed strong phylogenetic signal based on randomization tests and were significantly correlated, while latitude was not correlated with reproductive mode. This study partially supports the cold-climate model, but also suggests that altitude either may be a better predictor of cold temperatures or may be a surrogate for other selective factors important in the evolution of viviparity.